Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations IDS on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Motor problem - loss of torque, slow acceleration with open shaft, noisy 1

Status
Not open for further replies.

ScottyUK

Electrical
May 21, 2003
12,915
I got asked to give a second opinion on a motor which we have at our overhaul contractor's works. The machine is a form-wound 3300V 250kW type on a 355 frame. It suffered a failed DE bearing which cooked itself, hence it was removed from site and sent for overhaul. The stator was untouched during the rebuild and we don't believe there has ever been a rewind, so we are reasonably confident that the stator connections are correct. The physical construction means that inadvertently swapping polarity of a winding could only occur during a rewind. The stator passed a Baker test.

Following rebuild the motor is noisy with a continuous low-frequency noise, not a growl or buzz, more of a rumble. The motor accelerates very slowly on an open shaft test where there is only windage and bearing friction to overcome. The rotor is mechanically free and coasts for a long period once moving. On this basis we think there is a gross lack of torque during acceleration rather than any mechanical drag. If the bearings were dragging to the point of stalling a 250kW machine they would be torn to bits, and this isn't happening. During a start attempt the motor crawled up to speed over about a minute or so, and the peak current captured by the clamp meter was just over running current, i.e. abnormally low for a DOL start.

The motor load is a centrifugal pump which currently presents very little load due to process conditions requiring very low flow, so it is entirely possible that it has been running in this state for a while without causing any apparent problem at the load..

My opinion from the symnptoms is that there is something wrong with the rotor rather than the stator. My current theory is that there is a fracture in the squirrel cage end-ring itself, and that some current is passing into a relatively high resistance steel retaining ring located inside the copper end-ring. As a result the currents in the squirrel cage are much smaller than expected, causing loss of torque and keeping the line current low. There a few electrical burns which indicate passage of current and perhaps light arcing. The brazes where the bars meet the ring, and the ring itself, are obscured by a heavy film of varnish and overspray. There are no visible cracks, but I'm getting it cleaned down to bare metal in prep for dye penentrant and/or ultrasonic NDT on the end ring and the brazing.

Has anyone seen a similar set of symptoms before? Am I on the right track, or am I totally off into the weeds? Any thoughts on other possible sources of the problem?

I am not certain whether the copper ring was cast and machined, or hewn out of solid plate: can anyone with experience of manufacturing this type of machine offer an opinion which is most likely? It may be of significance to the NDT guys, plus it will influence whether repair is viable or not.

 
Replies continue below

Recommended for you

I should have also mention that single phase test is used by our local shops in my part of the USA.

=====================================
(2B)+(2B)' ?
 
Epete,
I am from Eastern Europe and I guess that's zlatkodo is from the same area .The principle to give on the motor small percent of the voltage and gently turning is very easy to perform .You should not have a special preparation for this test because it might be the start of open shaft testing after reparation of electric motor. I don*t have some references for this test in English but we suggest that you should try to perform . Mainly worker who hold the shaft gives an indication to colleague who raises the voltage until he can control the shaft.There are not safety consideration because the torque of motors is very small and it is easily control shaft of the motor . Sometimes the start electric motors with controlled voltage when passing cut down torque is moment when can be seen variation of current on ammeters and for experienced people in our repair shop can be a sign of damage to the rotor .
Good luck
 
With the motor running open shaft at full speed, you will see the current swinging very noticably at a low frequency (slipfrequency) if you have a damaged rotor bar.
I have observed currents swinging on a standard moving iron meter with the current swinging at 40+ %.

Otherwise, I favour the single phase approach with a low voltage applied and the rotor slowly turned. Easy and safe to do.
Best regards,
Mark

Mark Empson
Advanced Motor Control Ltd
 
With the motor running open shaft at full speed, you will see the current swinging very noticably at a low frequency (slipfrequency) if you have a damaged rotor bar.
I have observed currents swinging on a standard moving iron meter with the current swinging at 40+ %.
Hmm. Well at no load, the vast majority of the stator current is magnetizing current, which is unaffected by the rotor bar defect. There is very little current in the rotor (only enough to produce torque to overcome losses) which is affected by the rotor bar defect.

With a large magnetizing component unaffected by rotor bar defect and a tiny load component affected by rotor bar defect, I don't think that looking for no-load current variation will be a very sensitive test.


=====================================
(2B)+(2B)' ?
 
And sleep frequency is almost zero.Motor must be loaded and at pump aggregates where the load is constant variation of current is obvious on old iron ammeter . I worked for several years Motor current Signature analysis and found over 50 problems on the rotor at low voltage and high voltage motors .Lastly I could on the basis of the ammeter to the cubicles and current variation to predict that the motor has a problem with the rotor bares.
Good luck .
 
A word of caution.

I have been working with electric motors and drives for more than 40 years now. And I have quite often been asked to go and diagnose motors with suspect rotors. I must have been checking at least 500 motors - probably more.

I have only found four (that I can remember) actual rotor bar or shorting ring failures. All other cases were false alarms with fluctuating load and unstable drive being the most common cause of ammeter swinging. The second most common cause is interference between motor frequency and grid frequency when the DC link capacitors have a high ESR or have failed.

BTW, it is SLIP, not SLEEP. And BARS not BARES. Sorry to be such a brute, but I think that you are better off knowing the right words.

I agree that the motor must be loaded for the rotor bar damage to show up significantly on an ammeter. Low voltage and turning the rotor while observing current is a simple and safe way to diagnose broken rotor bar and end ring problems.

Gunnar Englund
--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.
 
Hi Pete

While I agree with your theory, I was asked to look at a machine some years ago, on a star/delta starter (wye delta) that would not start in star, but did start in delta albeit at much reduced current. When the motor was running, the current was observed to swing significantly at around 2 - 4 second cycle time.
The motor was dismantled and there were a number of broken rotor bars that had been arcing.
When these bars were repaired, the motor performed correctly.

Gunnar, I agree, very few problems are actually due to rotor bars, but I did see one large motor that had frequent ring problems.
This motor had been poorly manufactured with the rings breaking away from the bars after a relatively short time.
The repairer decided that the problem was related to too much overhang of the bars, so shortened them considerably before re welding the rings back on. There was now little room for expansion and cracked welds became a frequent problem. This motor was also on a star delta starter so subject to high transients during start.
Low voltage and turning the rotor while observing current is a simple and safe way to diagnose broken rotor bar and end ring problems.
Works for me!!

Mark Empson
Advanced Motor Control Ltd
 
Thanks Mark.

My comments (hm, well) expressed surprise because I was surprised to hear what you said.

Now thinking some more and seeing your explanation - if I had to guess I'd say that was a very severe rotor defect, based on
1 - reduced current noticeable during start. Looking at equivalent circuit, the effect of increasing R2 is to decrease both current and torque during the start.
2 - 2-4 second pole pass frequency even at open-shaft/no-load conditions. That implies the motor is really struggling just to make up for normal losses.

So maybe the friction losses were enough loading to create a significant load component in your case. (would be interesting to know how the average measured no-load current compared to fla and to expected no-load current).

At any rate I'll stick with my comment that it's not "sensitive". Meaning it is not a good way to ferret out a hard-to-find problem. But it's interesting to know maybe no-load current oscillation will show up for a really severe problem.

My comments (like a lot recently) are a little off-topic to Scotty's problem, but I still think it's interesting.

Going back to Scotty's op
Scotty said:
During a start attempt the motor crawled up to speed over about a minute or so, and the peak current captured by the clamp meter was just over running current, i.e. abnormally low for a DOL start.
Abnormally low current. Seems consistent with what you (Mark) reported.


=====================================
(2B)+(2B)' ?
 
Clarification:
(electricpete said:
So maybe the friction losses were enough loading to create a significant load component in your case.
should have been
(electricpete said:
So maybe the friction losses were enough loading to create a significant load torque component which was significant compared to the drastically reduced torque from severe defect in your case.


=====================================
(2B)+(2B)' ?
 
maybe both were right. The losses created a:
- load torque significant compared to reduced torque capability
- load current component significant compared to magnetizing current

=====================================
(2B)+(2B)' ?
 
Hi Pete

Yes, if you consider the rotor with essentially all bars open circuit, or close to it and just one relatively high impedance loop still in circuit (2 bars) then the lowest impedance could appear as two poles on the rotor and as that rotates at just below the synch speed, that low impedance will slowly rotate through the phases.

My suggestion was in response to Scotty's description which was very low start current and of course that reflects in an even lower start torque, If the bar to ring bonds are all badly cracked, then it is possible to have essentially a high impedance path primarily through two bars only.
It will be interesting to hear what they find.


Mark Empson
Advanced Motor Control Ltd
 
That's an extreme rotor failure. Never come across anything like it.

But, when you think about it, it all sounds very plausible.

Gunnar Englund
--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor